Association Between Vascular ¹⁸F-Fluorodeoxyglucose Uptake at Diagnosis and Change in Aortic Dimensions in Giant Cell Arteritis

Background: Previous studies have shown that patients with giant cell arteritis (GCA) who have vascular 18F-fluorodeoxyglucose (FDG) uptake at diagnosis are at increased risk for thoracic aortic complications. Objective: To measure the association between vascular FDG uptake at diagnosis and the change in aortic dimensions. Design: Prospective cohort study. Setting: University Hospitals Leuven. Patients: 106 patients with GCA and FDG positron emission tomography (PET) imaging 3 days or less after initiation of glucocorticoids. Measurements: Patients had PET and computed tomography (CT) imaging at diagnosis and CT imaging yearly for a maximum of 10 years. The PET scans were scored 0 to 3 in 7 vascular areas and summed to a total vascular score (TVS). The PET scan results were positive when FDG uptake was grade 2 or greater in any large vessel. The association between vascular FDG uptake and aortic dimensions was estimated by linear mixed-effects models with random intercept and slope. Results: When compared with patients with a negative PET scan result, those with a positive scan result had a greater increase in the diameter of the ascending aorta (difference in 5-year progression, 1.58 mm [95% CI, 0.41 to 2.74 mm]), the diameter of the descending aorta (1.32 mm [CI, 0.38 to 2.26 mm]), and the volume of the thoracic aorta (20.5 cm³ [CI, 4.5 to 36.5 cm³]). These thoracic aortic dimensions were also positively associated with TVS. Patients with a positive PET scan result had a higher risk for thoracic aortic aneurysms (adjusted hazard ratio, 10.21 [CI, 1.25 to 83.3]). Limitation: The lengthy inclusion and follow-up period resulted in missing data and the use of different PET machines. Conclusion: Higher TVS was associated with greater yearly increase in thoracic aortic dimensions. Performing PET imaging at diagnosis may help to estimate the risk for aortic aneurysm formation. Primary Funding Source: None.